Categorization of regional and foreign accent in 5- to 7-year-old British children

Abstract

This study examines children's ability to detect accent-related information in connected speech. British English children aged 5 and 7 years old were asked to discriminate between their home accent from an Irish accent or a French accent in a sentence categorization task. Using a preliminary accent rating task with adult listeners, it was first verified that the level of accentedness was similar across the two unfamiliar accents. Results showed that whereas the younger children group behaved just above chance level in this task, the 7-year-old group could reliably distinguish between these variations of their own language, but were significantly better at detecting the foreign accent than the regional accent. These results extend and replicate a previous study (Girard, Floccia, & Goslin, 2008) in which it was found that 5-year-old French children could detect a foreign accent better than a regional accent. The factors underlying the relative lack of awareness for a regional accent as opposed to a foreign accent in childhood are discussed, especially the amount of exposure, the learnability of both types of accents, and a possible difference in the amount of vowels versus consonants variability, for which acoustic measures of vowel formants and plosives voice onset time are provided.

Effects of contextual support on preschoolers’ accented speech comprehension

“…This account of course presumes that young children can detect (explicitly or implicitly) that a speaker has an accent, a task that other studies suggest is difficult for 5-year-olds (Floccia, Butler, Girard, & Goslin, 2009;Girard, Floccia, & Goslin, 2008;Wagner, Clopper, & Pate, 2014; though see Kinzler, Dupoux, & Spelke, 2007, for a different account).…”

Effects of Language Background on the Development of Sociolinguistic Awareness: The Perception of Accent Variation in Monolingual and Multilingual 5- to 7-Year-Old Children

“…One explanation for these findings is that at least initially, children's representations are organized along a quantitatively defined continuum, anchored by their "home" accent, which forms the core set of their experiences, with regional accents treated as intermediate between the home and foreign accent (Wagner et al, 2014). With more experience, older children are able to reliably distinguish their own from another unfamiliar regional variety (Floccia et al, 2009) and make judgements in line with general social stereotypes. For example, children aged 9-10 years in Illinois and Tennessee, USA, judged northern-accented American English speakers to be more intelligent than those with southern-accented American English (Kinzler & DeJesus, 2013).…”

“…Few studies have investigated this question, and the results are mixed. For example, Jeffries (2016) found that monolingual English children aged 3-4 years, growing up in Yorkshire but with a parent from outside Yorkshire (i.e., who spoke with a different accent from that of the community), were better able to group talkers according to ac-cent when the target variables contained different rather than the same phonemes (see also Floccia et al, 2009). Interestingly, children with parents from outside Yorkshire were also better able to recognize a familiar talker (in this case, a familiar teacher) using phonetic accent features.…”

“…To assess whether or not children were able to detect the unfamiliar/untrained accent using a bias-free measure, separate measures of sensitivity (A') and bias (B" D ) were computed using signal detection analysis (see Girard et al, 2008;Floccia et al, 2009). Sensitivity (A') was used rather than d' (Grier, 1971) because it does not rely on the strict assumptions of normality and equal variance of signal and noise distributions.…”

“…Thus, in the home-regional condition, all children, regardless of language background, had a bias towards answering that the talker had the "home" accent, whilst in the regional-foreign condition, all children had a bias towards answering that the talker had the "foreign" accent. Previous research with monolingual children has shown that the ability to explicitly categorize talkers according to regional accent begins to emerge at around 5 years, and that initially they can only successfully categorize speakers when the acoustic-phonetic differences between the accents are maximized (Floccia et al, 2009;Girard et al, 2008;Wagner et al, 2014). This led researchers to hypothesize that children have gradient representations of accent variation, with their home accent, the one that is experienced the most, forming the core set of experiences and other varieties identified relative to this (Wagner et al, 2014).…”